Philemon Forson

AI/ML Engineer & Data Scientist

M.S. Business Analytics & AI, The University of Texas at Dallas · Dallas, TX

Building practical AI systems at the intersection of machine learning, software engineering, and data.

Experience

  1. Artificial Intelligence Engineer Intern

    Zof AI

    San Francisco, California — Remote

    February 2026 – May 2026

    Evaluated end-to-end product workflows, identifying usability issues, missing functionality, and potential failure points.

  2. Software Engineer

    Logiciel Limited

    Accra, Ghana

    November 2023 – January 2025

    Led an engineering team through the migration from Next.js 14 to Next.js 15 and adoption of Turbopack, improving local development performance, shortening feedback cycles, and saving 14 developer-hours per month.

  3. Data Engineer — Future Lab

    GRUNDFOS Ghana Limited

    Accra, Ghana

    August 2022 – May 2023

    Built Python ETL workflows to ingest customer and IoT data from APIs and internal systems, then clean, validate, transform, and prepare it for analytics, machine learning, and data science workflows.

  4. Data Analyst

    Foresight Consulting Ltd

    Manchester, United Kingdom

    February 2021 – June 2022

    Analyzed and cleaned business datasets using Python, pandas, and Excel, resolving missing values, duplicate records, and inconsistent formatting.

Projects

Traffic Intelligence Platform

Computer Vision · Live

End-to-end computer vision platform for vehicle and pedestrian detection, multi-object tracking, trajectory analysis, and traffic analytics. Includes real-time detection, directional counting, FastAPI inference, annotated playback, speed estimation, license-plate OCR, and VLM-based event analysis with semantic retrieval.

  • Python
  • PyTorch
  • OpenCV
  • YOLO
  • ByteTrack
  • FastAPI
  • Docker

Credex — Explainable Credit Risk Modeling Platform

AI / ML · Live

End-to-end credit risk platform built from 1.1M+ transaction records, combining predictive modeling, explainability, APIs, and a web interface. Addresses class imbalance and temporal feature engineering, benchmarks XGBoost against a deep neural network (AUC 0.945), and serves SHAP explanations through a Flask API and Next.js frontend.

  • Python
  • SQL
  • XGBoost
  • TensorFlow
  • SHAP
  • Flask
  • Next.js
  • Docker

Soccer Match LLM Analyst

GenAI / RAG · Live

Full-stack AI application for soccer match search and video discovery using RAG, semantic retrieval, ranking logic, and a Next.js/FastAPI architecture. Uses SentenceTransformers and ChromaDB with recency-aware ranking, home/away team interpretation, official-source prioritization, and automated filtering of simulations and low-quality content.

  • Python
  • Next.js
  • FastAPI
  • GPT-4
  • ChromaDB
  • SentenceTransformers
  • Docker

Get In Touch

I'm open to AI/ML, software engineering, and data-focused opportunities where I can contribute, keep learning, and build systems that create real value.

Email

simplelogin-newsletter.contour409@simplelogin.com

The best way to reach me about opportunities or projects.

GitHub

github.com/phil-forson

Code, experiments, and public repositories.

LinkedIn

linkedin.com/in/forson-philemon

Professional background and experience.

Location

Dallas, Texas

Based in Dallas, Texas.

Let's build something useful.

If you're working on something in AI/ML, software, or data and think I could help, I'd like to hear from you.